Spaces:
Running
Running
Commit
·
4022606
1
Parent(s):
647d063
Update app.py
Browse files
app.py
CHANGED
@@ -58,84 +58,96 @@ if st.button("Process"):
|
|
58 |
columns = [x.lower() for x in columns]
|
59 |
df.columns = columns
|
60 |
print(summarizer_option)
|
61 |
-
|
62 |
-
|
63 |
-
|
64 |
-
|
65 |
-
|
66 |
-
|
67 |
-
|
68 |
-
|
69 |
-
|
70 |
-
|
71 |
-
|
72 |
-
|
73 |
-
|
74 |
-
|
75 |
-
|
76 |
-
|
77 |
-
|
78 |
-
|
79 |
-
|
80 |
-
|
81 |
-
|
82 |
-
|
83 |
-
data=csv,
|
84 |
-
file_name=f"{summarizer_option}_df.csv",
|
85 |
-
mime="text/csv",
|
86 |
-
)
|
87 |
-
if summarizer_option == "t5-base":
|
88 |
-
model, tokenizer = load_t5()
|
89 |
-
text = df["text"].values.tolist()
|
90 |
-
summary = []
|
91 |
-
for x in stqdm(range(len(text))):
|
92 |
-
|
93 |
-
tokens_input = tokenizer.encode(
|
94 |
-
"summarize: " + text[x],
|
95 |
-
return_tensors="pt",
|
96 |
-
max_length=tokenizer.model_max_length,
|
97 |
-
truncation=True,
|
98 |
)
|
99 |
-
|
100 |
-
|
101 |
-
|
102 |
-
|
103 |
-
|
104 |
-
|
105 |
)
|
106 |
-
|
107 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
108 |
|
109 |
-
|
110 |
-
|
111 |
-
|
112 |
-
|
113 |
-
|
114 |
-
|
115 |
-
|
116 |
-
|
117 |
-
|
118 |
-
|
119 |
|
120 |
-
|
121 |
-
|
122 |
-
|
123 |
|
124 |
-
|
125 |
|
126 |
-
|
127 |
-
|
128 |
-
|
129 |
-
|
130 |
-
|
131 |
-
|
132 |
-
|
133 |
-
|
134 |
-
|
135 |
-
|
136 |
-
|
137 |
-
|
138 |
-
|
139 |
-
|
140 |
-
|
|
|
|
|
|
|
|
|
|
|
141 |
)
|
|
|
|
58 |
columns = [x.lower() for x in columns]
|
59 |
df.columns = columns
|
60 |
print(summarizer_option)
|
61 |
+
try:
|
62 |
+
|
63 |
+
if summarizer_option == "Custom trained on the dataset":
|
64 |
+
model = custom_model()
|
65 |
+
print(summarizer_option)
|
66 |
+
text = df["text"].values.tolist()
|
67 |
+
progress_text = "Summarization in progress. Please wait."
|
68 |
+
summary = []
|
69 |
+
|
70 |
+
for x in stqdm(range(len(text))):
|
71 |
+
try:
|
72 |
+
summary.append(
|
73 |
+
model(
|
74 |
+
f"summarize: {text[x]}",
|
75 |
+
max_length=50,
|
76 |
+
early_stopping=True,
|
77 |
+
)[0]["summary_text"]
|
78 |
+
)
|
79 |
+
except:
|
80 |
+
pass
|
81 |
+
output = pd.DataFrame(
|
82 |
+
{"text": df["text"].values.tolist(), "summary": summary}
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
83 |
)
|
84 |
+
csv = convert_df(output)
|
85 |
+
st.download_button(
|
86 |
+
label="Download data as CSV",
|
87 |
+
data=csv,
|
88 |
+
file_name=f"{summarizer_option}_df.csv",
|
89 |
+
mime="text/csv",
|
90 |
)
|
91 |
+
if summarizer_option == "t5-base":
|
92 |
+
model, tokenizer = load_t5()
|
93 |
+
text = df["text"].values.tolist()
|
94 |
+
summary = []
|
95 |
+
for x in stqdm(range(len(text))):
|
96 |
+
|
97 |
+
tokens_input = tokenizer.encode(
|
98 |
+
"summarize: " + text[x],
|
99 |
+
return_tensors="pt",
|
100 |
+
max_length=tokenizer.model_max_length,
|
101 |
+
truncation=True,
|
102 |
+
)
|
103 |
+
summary_ids = model.generate(
|
104 |
+
tokens_input,
|
105 |
+
min_length=80,
|
106 |
+
max_length=150,
|
107 |
+
length_penalty=20,
|
108 |
+
num_beams=2,
|
109 |
+
)
|
110 |
+
summary_gen = tokenizer.decode(
|
111 |
+
summary_ids[0], skip_special_tokens=True
|
112 |
+
)
|
113 |
+
summary.append(summary_gen)
|
114 |
|
115 |
+
output = pd.DataFrame(
|
116 |
+
{"text": df["text"].values.tolist(), "summary": summary}
|
117 |
+
)
|
118 |
+
csv = convert_df(output)
|
119 |
+
st.download_button(
|
120 |
+
label="Download data as CSV",
|
121 |
+
data=csv,
|
122 |
+
file_name=f"{summarizer_option}_df.csv",
|
123 |
+
mime="text/csv",
|
124 |
+
)
|
125 |
|
126 |
+
if summarizer_option == "t5-one-line-summary":
|
127 |
+
model = SimpleT5()
|
128 |
+
text = df["text"].values.tolist()
|
129 |
|
130 |
+
load_one_line_summarizer(model=model)
|
131 |
|
132 |
+
summary = []
|
133 |
+
for x in stqdm(range(len(text))):
|
134 |
+
try:
|
135 |
+
summary.append(model.predict(text[x])[0])
|
136 |
+
except:
|
137 |
+
pass
|
138 |
+
output = pd.DataFrame(
|
139 |
+
{"text": df["text"].values.tolist(), "summary": summary}
|
140 |
+
)
|
141 |
+
csv = convert_df(output)
|
142 |
+
st.download_button(
|
143 |
+
label="Download data as CSV",
|
144 |
+
data=csv,
|
145 |
+
file_name=f"{summarizer_option}_df.csv",
|
146 |
+
mime="text/csv",
|
147 |
+
)
|
148 |
+
except KeyError:
|
149 |
+
st.error(
|
150 |
+
"Please Make sure that your data must have a column named text",
|
151 |
+
icon="🚨",
|
152 |
)
|
153 |
+
st.info("Text column must have amazon reviews", icon="ℹ️")
|